📊 Copilot Token Usage Report2026-04-28

[github-actions]

Overview

Period: 2026-04-27 to 2026-04-28
Runs analyzed: 50 (49 had token data)
Total tokens: 9,287,181 across all workflows
Estimated total cost: $0.00 (cost tracking not yet enabled — tokens tracked via AWF api-proxy)
Total Actions minutes: 221 min

All runs used model: claude-sonnet-4.6 (via Copilot provider)

Workflow Summary

Workflow	Runs	Total Tokens	Avg Tokens/Run	Turns	Action Min
Build Test Suite	8	3,078K	385K	7	41
Smoke Services	8	1,760K	220K	8	50
Smoke Copilot	9	1,033K	115K	9	30
Smoke Copilot BYOK	9	1,013K	113K	9	29
Firewall Issue Dispatcher	3	861K	287K	3	13
Issue Duplication Detector	3	582K	194K	5	11
Smoke Chroot	7	476K	68K	7	35
Daily Claude Token Usage Analyzer	1	298K	298K	1	8
Dependency Security Monitor	1	180K	180K	1	3
Daily Copilot Token Usage Analyzer	1	—	—	—	1

🔍 Optimization Opportunities

1. Build Test Suite — 385K tokens/run avg, 83 total API requests across 8 runs

   • High token volume; heavy cache usage (2.7M cache reads vs 3.0M input) is good — cache efficiency ~47%
   • 1 failure in 8 runs (12.5% error rate) — worth investigating

2. Smoke Services — 220K tokens/run avg, 2 cancelled runs

   • 2 cancellations out of 8 runs (25%) — may indicate timeout or resource contention issues
   • 2 errors in 8 runs (25% error rate); investigate reliability

3. Smoke Copilot BYOK — 113K tokens/run avg, 2 errors (1 failure, 1 cancelled)

   • 1 outright failure (run 25028438866) — investigate root cause
   • Error rate 22%; compare with Smoke Copilot (0 errors) to identify BYOK-specific issues

4. Issue Duplication Detector — 3 runs but 5 turns (avg 1.7 turns/run)

   • Higher-than-expected turns for a duplication check; consider pre-filtering to reduce agentic overhead

Per-Workflow Details

Build Test Suite

• Runs: 8 — 25032270162 ✅ 25032200418 ✅ 25032156274 ✅ 25030177629 ✅ 25029951361 ❌ 25030026673 ✅ 25007989181 ✅ 25008506453 ✅
• Total tokens: 3,078K (avg 385K/run)
• Effective tokens: 3,475K (cache read: 2,721K, cache efficiency ~47%)
• Model: claude-sonnet-4.6 — 3,036K input, 41K output, 83 requests
• Action minutes: 41 min total
• Error rate: 1/8 runs (12.5%)

Smoke Services

• Runs: 8 — 25032270125 ✅ 25032200426 ✅ 25032156277 🚫 25030177625 ✅ 25030026715 🚫 25029951356 ✅ 25028395483 ✅ 25008506604 ✅
• Total tokens: 1,760K (avg 220K/run)
• Effective tokens: 1,939K (cache read: 1,551K, cache efficiency ~49%)
• Model: claude-sonnet-4.6 — 1,752K input, 7.8K output, 63 requests
• Action minutes: 50 min total
• Error rate: 2/8 runs (2 cancelled, 25%)

Smoke Copilot

• Runs: 9 — all successful ✅
• Total tokens: 1,033K (avg 115K/run)
• Effective tokens: 1,142K (cache read: 804K, cache efficiency ~44%)
• Model: claude-sonnet-4.6 — 1,024K input, 9.5K output, 40 requests
• Action minutes: 30 min total
• Error rate: 0/9 runs (0%)

Smoke Copilot BYOK

• Runs: 9 — 25032270173 ✅ 25032200440 ✅ 25032156276 🚫 25030177641 ✅ 25030026655 ✅ 25029951319 ✅ 25028438866 ❌ 25008506557 ✅ 25007989173 ✅
• Total tokens: 1,013K (avg 113K/run)
• Effective tokens: 1,119K (cache read: 772K, cache efficiency ~43%)
• Model: claude-sonnet-4.6 — 1,004K input, 9.4K output, 39 requests
• Action minutes: 29 min total
• Error rate: 2/9 runs (22%)

Firewall Issue Dispatcher

• Runs: 3 — all successful ✅
• Total tokens: 861K (avg 287K/run)
• Effective tokens: 952K (cache read: 723K, cache efficiency ~46%)
• Model: claude-sonnet-4.6 — 855K input, 6.1K output, 25 requests
• Action minutes: 13 min total
• Error rate: 0/3 runs (0%)

Issue Duplication Detector

• Runs: 3 — all successful ✅
• Total tokens: 582K (avg 194K/run)
• Effective tokens: 1,036K — high effective vs raw ratio; 5 turns across 3 runs (avg 1.7/run)
• Model: claude-sonnet-4.6 — 942K input, 3.5K output, 31 requests
• Action minutes: 11 min total
• Error rate: 0/3 runs (0%)

Smoke Chroot

• Runs: 7 — 25032270144 ✅ 25032200448 🚫 25032156287 ✅ 25030177626 ✅ 25030026710 ✅ 25029951357 ✅ 25008506531 ✅
• Total tokens: 476K (avg 68K/run)
• Effective tokens: 522K
• Model: claude-sonnet-4.6 — 473K input, 3.8K output, 21 requests
• Action minutes: 35 min total
• Error rate: 1/7 runs (1 cancelled, 14%)

Daily Claude Token Usage Analyzer

• Runs: 1 — 25017103024 ✅
• Total tokens: 298K
• Model: claude-sonnet-4.6 — 293K input, 4.7K output, 8 requests (252K cache reads)
• Action minutes: 8 min

Dependency Security Monitor

• Runs: 1 — 25025311549 ✅
• Total tokens: 180K
• Model: claude-sonnet-4.6 — 179K input, 1.1K output, 5 requests (133K cache reads)
• Action minutes: 3 min

Workflows Without Token Data

Runs where token_usage was null or 0 — these may not have completed or may not have the api-proxy enabled:

• Daily Copilot Token Usage Analyzer — run 25048605390 (status: in_progress at data collection time)

Historical Trend

Comparing to previous report (issue #2233, period 2026-04-26 to 2026-04-27):

Metric	Previous	Current	Change
Total runs	28	50	+79%
Total tokens	8,689K	9,287K	+7%
Avg tokens/run	310K	186K	-40% ↓
Action minutes	117	221	+89%
Total errors	—	6	—

More runs this period (+79%) with significantly lower avg tokens per run (-40%), indicating more lightweight workflow runs (smoke tests, chroot) rather than one-off heavy analysis workflows. The Build Test Suite remains the dominant consumer at 3,078K tokens total, comparable to yesterday's 853K across only 2 runs — today's 8 runs show it's now running more frequently.

Previous Report

Issue #2233 — 📊 Copilot Token Usage Report 2026-04-27

References:

• §25048605390 — Current workflow run

Generated by Daily Copilot Token Usage Analyzer · ● 363.3K · ◷

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #2289

View newer issue

---

This action was performed automatically by the Daily Copilot Token Usage Analyzer workflow.